Difference between revisions of "cpp/algorithm/ranges/minmax"
From cppreference.com
Andreas Krug (Talk | contribs) m (fmt, headers sorted) |
Andreas Krug (Talk | contribs) m (fmt, @-@ -> @,@) |
||
| Line 46: | Line 46: | ||
{{par|r|a non-empty range of values to compare}} | {{par|r|a non-empty range of values to compare}} | ||
{{par|comp|comparison to apply to the projected elements}} | {{par|comp|comparison to apply to the projected elements}} | ||
| − | {{par|proj|projection to apply to the elements | + | {{par|proj|projection to apply to the elements}} |
{{par end}} | {{par end}} | ||
| Line 52: | Line 52: | ||
@1@ {{c|{b, a} }} if, according to their respective projected value, {{c|b}} is smaller than {{c|a}}; otherwise it returns {{c|{a, b} }}. | @1@ {{c|{b, a} }} if, according to their respective projected value, {{c|b}} is smaller than {{c|a}}; otherwise it returns {{c|{a, b} }}. | ||
| − | @2 | + | @2,3@ {{c|{s, l} }}, where {{tt|s}} and {{tt|l}} are respectively the smallest and largest values in {{c|r}}, according to their projected value. If several values are equivalent to the smallest and largest, returns the leftmost smallest value, and the rightmost largest value. If the range is empty (as determined by {{c|ranges::distance(r)}}), the behavior is undefined. |
===Complexity=== | ===Complexity=== | ||
@1@ Exactly one comparison and two applications of the projection. | @1@ Exactly one comparison and two applications of the projection. | ||
| − | @2 | + | @2,3@ At most {{c|3 / 2 * ranges::distance(r)}} comparisons and twice as many applications of the projection. |
===Possible implementation=== | ===Possible implementation=== | ||
| Line 127: | Line 127: | ||
namespace ranges = std::ranges; | namespace ranges = std::ranges; | ||
| − | constexpr std::array v {3, 1, 4, 1, 5, 9, 2, 6, 5}; | + | constexpr std::array v{3, 1, 4, 1, 5, 9, 2, 6, 5}; |
std::random_device rd; | std::random_device rd; | ||
| Line 148: | Line 148: | ||
std::cout << "smallest: " << min << ", " << "largest: " << max << '\n'; | std::cout << "smallest: " << min << ", " << "largest: " << max << '\n'; | ||
} | } | ||
| − | |p=true|output= | + | |p=true |
| + | |output= | ||
v[3:9]: 1 5 9 2 6 5 | v[3:9]: 1 5 9 2 6 5 | ||
smallest: 1, largest: 9 | smallest: 1, largest: 9 | ||
Latest revision as of 10:34, 28 August 2023
| Defined in header <algorithm>
|
||
| Call signature |
||
| template< class T, class Proj = std::identity, std::indirect_strict_weak_order< |
(1) | (since C++20) |
| template< std::copyable T, class Proj = std::identity, std::indirect_strict_weak_order< |
(2) | (since C++20) |
| template< ranges::input_range R, class Proj = std::identity, std::indirect_strict_weak_order< |
(3) | (since C++20) |
| Helper types |
||
| template< class T > using minmax_result = ranges::min_max_result<T>; |
(4) | (since C++20) |
Returns the smallest and the greatest of the given projected values.
1) Returns references to the smaller and the greater of a and b.
2) Returns the smallest and the greatest of the values in the initializer list r.
3) Returns the smallest and the greatest of the values in the range r.
The function-like entities described on this page are niebloids, that is:
- Explicit template argument lists cannot be specified when calling any of them.
- None of them are visible to argument-dependent lookup.
- When any of them are found by normal unqualified lookup as the name to the left of the function-call operator, argument-dependent lookup is inhibited.
In practice, they may be implemented as function objects, or with special compiler extensions.
Contents |
[edit] Parameters
| a, b | - | the values to compare |
| r | - | a non-empty range of values to compare |
| comp | - | comparison to apply to the projected elements |
| proj | - | projection to apply to the elements |
[edit] Return value
1) {b, a} if, according to their respective projected value, b is smaller than a; otherwise it returns {a, b}.
2,3) {s, l}, where
s and l are respectively the smallest and largest values in r, according to their projected value. If several values are equivalent to the smallest and largest, returns the leftmost smallest value, and the rightmost largest value. If the range is empty (as determined by ranges::distance(r)), the behavior is undefined.[edit] Complexity
1) Exactly one comparison and two applications of the projection.
2,3) At most 3 / 2 * ranges::distance(r) comparisons and twice as many applications of the projection.
[edit] Possible implementation
struct minmax_fn { template<class T, class Proj = std::identity, std::indirect_strict_weak_order< std::projected<const T*, Proj>> Comp = ranges::less> constexpr ranges::minmax_result<const T&> operator()(const T& a, const T& b, Comp comp = {}, Proj proj = {}) const { if (std::invoke(comp, std::invoke(proj, b), std::invoke(proj, a))) return {b, a}; return {a, b}; } template<std::copyable T, class Proj = std::identity, std::indirect_strict_weak_order< std::projected<const T*, Proj>> Comp = ranges::less> constexpr ranges::minmax_result<T> operator()(std::initializer_list<T> r, Comp comp = {}, Proj proj = {}) const { auto result = ranges::minmax_element(r, std::ref(comp), std::ref(proj)); return {*result.min, *result.max}; } template<ranges::input_range R, class Proj = std::identity, std::indirect_strict_weak_order< std::projected<ranges::iterator_t<R>, Proj>> Comp = ranges::less> requires std::indirectly_copyable_storable<ranges::iterator_t<R>, ranges::range_value_t<R>*> constexpr ranges::minmax_result<ranges::range_value_t<R>> operator()(R&& r, Comp comp = {}, Proj proj = {}) const { auto result = ranges::minmax_element(r, std::ref(comp), std::ref(proj)); return {std::move(*result.min), std::move(*result.max)}; } }; inline constexpr minmax_fn minmax; |
[edit] Notes
For overload (1), if one of the parameters is a temporary, the reference returned becomes a dangling reference at the end of the full expression that contains the call to minmax:
int n = 1; auto p = std::ranges::minmax(n, n + 1); int m = p.min; // ok int x = p.max; // undefined behavior // Note that structured bindings have the same issue auto [mm, xx] = std::ranges::minmax(n, n + 1); xx; // undefined behavior
[edit] Example
Run this code
#include <algorithm> #include <array> #include <iostream> #include <random> int main() { namespace ranges = std::ranges; constexpr std::array v{3, 1, 4, 1, 5, 9, 2, 6, 5}; std::random_device rd; std::mt19937_64 generator(rd()); std::uniform_int_distribution<> distribution(0, ranges::distance(v)); // [0..9] // auto bounds = ranges::minmax(distribution(generator), distribution(generator)); // UB: dangling references: bounds.min and bounds.max have the type `const int&`. const int x1 = distribution(generator); const int x2 = distribution(generator); auto bounds = ranges::minmax(x1, x2); // OK: got references to lvalues x1 and x2 std::cout << "v[" << bounds.min << ":" << bounds.max << "]: "; for (int i = bounds.min; i < bounds.max; ++i) std::cout << v[i] << ' '; std::cout << '\n'; auto [min, max] = ranges::minmax(v); std::cout << "smallest: " << min << ", " << "largest: " << max << '\n'; }
Possible output:
v[3:9]: 1 5 9 2 6 5 smallest: 1, largest: 9
[edit] See also
| (C++20) |
returns the smaller of the given values (niebloid) |
| (C++20) |
returns the greater of the given values (niebloid) |
| (C++20) |
returns the smallest and the largest elements in a range (niebloid) |
| (C++20) |
clamps a value between a pair of boundary values (niebloid) |
| (C++11) |
returns the smaller and larger of two elements (function template) |
